Follow the latest coverage, related explainers and connected technology stories.
Meta has decided to suspend proposed updates to its commercial terms of service and launch an extensive consultation process with publishers to address concerns related to the operation of web crawlers.
Apple plans to launch its first smart glasses in 2027, focusing on integrating the Siri voice assistant and iPhone apps rather than relying on augmented reality.
The list of public places banning the use of camera-equipped smart glasses is expanding worldwide, driven by growing concerns regarding privacy, illegal surveillance, and the protection of individual rights.
Meta has launched official support for the Disney 3D movie library via the Disney+ app on Quest 3 and 3S headsets, providing users with an immersive cinematic viewing experience at home.
Judicial documents have revealed internal discussions at Meta in 2023 aimed at undermining the credibility of researcher Jonathan Haidt, sparking a wave of widespread criticism from academics and researchers.
Meta has launched a new feature for its smart agent Muse that enables it to make calls on behalf of users with US companies, but exclusively for paid plan subscribers.
Meta has announced Muse Ember, its smallest AI model, specifically designed to execute repetitive tasks for digital agents at a very low cost with multimodal capabilities.
Meta has announced the global expansion of its AI-powered age estimation technology for Instagram users, aiming to automatically subject teenagers to account protection policies.
Meta has begun rolling out real-time voice conversation features and expressive avatars in the Muse application, allowing users to have natural and seamless dialogues with artificial intelligence across various platforms.
Anthropic has announced its new smart agent, Claude Duet, which operates proactively via a virtual machine connected to accounts and applications, and is currently available to subscribers of Pro plans and above via a waitlist in the United States.
A US court in New Mexico has imposed a $25 billion fine on Meta in a case involving user privacy violations and data leaks linked to the Cambridge Analytica scandal.
Meta has reactivated the item sharing feature in its Muse AI assistant while replacing all old links, following the discovery by security researchers of a flaw that allowed private internal documents to be found online.
Meta has removed explicit phrases referring to AI training on news websites from its new business terms, yet legal experts confirm that the amendments still leave loopholes allowing for collection and training.
AI agents from OpenAI, Meta, Google, xAI, and other companies face growing restrictions that prevent them from carrying out tasks such as shopping, booking, and accessing accounts. This reveals that agents’ ability to perform tasks online depends on website approval and compatibility with anti-bot systems, not solely on their software capabilities.
Meta has launched the open-weight Muse Spark Open artificial intelligence model, designed for coding and long-running agentic tasks, which can run locally on two Mac Studio devices equipped with M5 Ultra processors and a combined 1 terabyte of unified memory. Although its weights are available for download, the model is subject to Meta’s license, which imposes commercial restrictions on certain uses.
Meta announced the Muse Wonder and Muse Wonder Full series, two models designed for complex tasks and intelligent agents, with improved abilities to remember information and continue working after setbacks. They are now available to developers through the Meta Model API and Muse Code, and will later reach Muse app users, beginning with subscribers.
OpenAI and Meta are competing to offer AI agents that promise better protection for users’ data, but Muse’s security issues and its data-collection behaviors reveal the gap between marketing and practice. These systems’ ability to access sensitive information, along with the limits of user control, remain decisive factors before users can trust them.
Butterfly Effect, the owner of the Manus artificial intelligence laboratory, has raised more than $500 million in its first funding round since Meta’s $2 billion acquisition deal for the company was canceled. The company plans to continue hiring globally without disclosing its current valuation.
Google has made SynthID.com available for checking images, video, and audio for invisible watermarks used by Google and its partners, including OpenAI and Nvidia, with a daily limit of approximately 10 checks per user. However, the tool does not detect all AI-generated content, as it depends on model providers that use SynthID and does not prevent watermark removal.
Meta launched the open-source Muse Gadgets project, providing firmware and an SDK for Linux to build physical devices that connect to the Muse AI agent. This includes prototypes such as e-ink displays and units that connect to televisions, along with the internally developed Home Link device, which has limited distribution.